]>
granicus.if.org Git - libvpx/log
Jingning Han [Thu, 27 Feb 2014 20:00:24 +0000 (12:00 -0800)]
Remove arguments not in use in non-RD decision
This commit removes a few arguments that are not used in the
non-RD mode decision.
Change-Id: I0d878bb5a7dba1b34fe0b64afe55bf11b1a9a2d7
Jingning Han [Thu, 27 Feb 2014 19:41:04 +0000 (11:41 -0800)]
Cosmetic change mbmi->ref_frame to refs
In handle_inter_mode, the reference frames are set in refs buffer.
One can use refs buffer directly to avoid redundant fetch.
Change-Id: I811d408cae52dcd5e053dd4bfe69550eb6a2ff56
Jingning Han [Thu, 27 Feb 2014 19:26:39 +0000 (11:26 -0800)]
Assign motion vector invalid value in intra mode
In non-RD mode decision, assign motion vector INVALID_MV when intra
prediction mode is selected.
Change-Id: I8f6ec39b71e755758f0f698074a2c17d934622ea
Jingning Han [Thu, 27 Feb 2014 19:04:15 +0000 (11:04 -0800)]
Merge "Run sub-pixel motion search for NEWMV mode test"
Minghai Shang [Thu, 27 Feb 2014 18:51:26 +0000 (10:51 -0800)]
Merge "[svc] Add target bitrate settings for each layers."
Minghai Shang [Thu, 27 Feb 2014 18:50:02 +0000 (10:50 -0800)]
Merge "Change for adding QP settings for key frames"
Dmitry Kovalev [Thu, 27 Feb 2014 17:55:46 +0000 (09:55 -0800)]
Merge "Moving common code into vp9_get_entropy_contexts()."
Dmitry Kovalev [Thu, 27 Feb 2014 17:55:18 +0000 (09:55 -0800)]
Merge "Replacing int_mv with MV in single_motion_search()."
Dmitry Kovalev [Thu, 27 Feb 2014 17:54:50 +0000 (09:54 -0800)]
Merge "Removing vp9_systemdependent.c."
Deb Mukherjee [Thu, 27 Feb 2014 17:52:01 +0000 (09:52 -0800)]
Merge "Refines variance based partitioning search"
Deb Mukherjee [Wed, 26 Feb 2014 01:22:34 +0000 (17:22 -0800)]
Refines variance based partitioning search
Instead of using source variance, this patch uses variance of the
frame difference between the source and the current frame to make
fixed size partition decisions. Also disables adjusting partitioning
if variance based or fixed size partitioning is used.
The latter change improves the speed substantially for speed 6, so
that speed 7 is now less than 3x the speed of speed 6. But speed
6 is 48% better in psnr on the rtc set compared to speed 7.
As compared to speed 5,
speed 6 is -37% in psnr at about 2.5x the speed,
speed 7 is -55% in psnr at about 7x the speed.
Change-Id: If61d80431d3e04ed304ac05832e773cdb2c0a578
Yaowu Xu [Thu, 27 Feb 2014 16:56:32 +0000 (08:56 -0800)]
Merge "vp9_encodemv.h: fix function prototype"
James Zern [Thu, 27 Feb 2014 07:53:24 +0000 (23:53 -0800)]
vp9_encodemv.h: fix function prototype
mismatch was introduced here:
abe21e4 Removing redundant vp9_clear_system_state() call.
Change-Id: I66f945c1ce385dddfdefa9f517f295dbdd01d286
James Zern [Thu, 27 Feb 2014 07:45:09 +0000 (23:45 -0800)]
Merge "remove ioc problem with rdcost calculation"
James Zern [Thu, 27 Feb 2014 07:41:57 +0000 (23:41 -0800)]
Merge "gen_msvs_vcxproj.sh: Avoid object name collisions."
James Zern [Thu, 27 Feb 2014 07:34:40 +0000 (23:34 -0800)]
Merge changes I03c1ba21,I6d8ac50c,Id0a2d9ef,I3db3acbd,I028bdec9,I26f8724b,I3795d78b,I6bb7ffe8
* changes:
variance_test: add NEON functions
subtract_test: add NEON functions
sixtap_predict_test: add NEON functions
partial_idct_test: add NEON functions
fdct8x8_test: add NEON functions
fdct4x4_test: add NEON functions
dct32x32_test: add NEON functions
dct16x16_test: add NEON functions
James Zern [Wed, 26 Feb 2014 07:11:49 +0000 (23:11 -0800)]
variance_test: add NEON functions
note not all functions have NEON implementations:
- variance4x4_neon
Change-Id: I03c1ba21f3b02aa2482d7ca8feedc3ef74b5947f
James Zern [Wed, 26 Feb 2014 07:11:49 +0000 (23:11 -0800)]
subtract_test: add NEON functions
Change-Id: I6d8ac50ce842749a2a643d8173bbd301cc284236
James Zern [Wed, 26 Feb 2014 07:11:49 +0000 (23:11 -0800)]
sixtap_predict_test: add NEON functions
Change-Id: Id0a2d9ef9636c84ee988dabecd5d4dc2b8641944
James Zern [Wed, 26 Feb 2014 07:11:49 +0000 (23:11 -0800)]
partial_idct_test: add NEON functions
note not all functions have NEON implementations:
- vp9_idct32x32_34_add
Change-Id: I3db3acbd5b33839aabd1beadc6e0742ec50b1072
James Zern [Wed, 26 Feb 2014 07:11:49 +0000 (23:11 -0800)]
fdct8x8_test: add NEON functions
FwdTrans8x8HT is disabled as the tests currently fail.
note not all functions have NEON implementations:
- fdct8x8/fht8x8
Change-Id: I028bdec9a21eaaee2c5865470ab179aac403540e
James Zern [Wed, 26 Feb 2014 07:11:49 +0000 (23:11 -0800)]
fdct4x4_test: add NEON functions
Trans4x4HT is disabled as the tests currently fail.
note not all functions have NEON implementations:
- fdct4x4/fht4x4
Change-Id: I26f8724bf2a9ea01d59205a1c57119ed25d043bc
James Zern [Wed, 26 Feb 2014 07:11:49 +0000 (23:11 -0800)]
dct32x32_test: add NEON functions
note not all functions have NEON implementations:
- vp9_fdct32x32/vp9_fdct32x32_rd
Change-Id: I3795d78bf215e41ab8801c2c37cf633f16931f9d
Alex Converse [Thu, 27 Feb 2014 02:50:00 +0000 (18:50 -0800)]
Merge "Don't let GOLDEN_FRAME turn off LAST_FRAME as a reference when GOLDEN_FRAME is disabled."
Dmitry Kovalev [Thu, 27 Feb 2014 02:38:26 +0000 (18:38 -0800)]
Merge "Adding vpx_sse_to_psnr() function."
Dmitry Kovalev [Thu, 27 Feb 2014 02:37:21 +0000 (18:37 -0800)]
Merge "Cleaning up cost_coeffs()."
Dmitry Kovalev [Thu, 27 Feb 2014 02:07:23 +0000 (18:07 -0800)]
Removing vp9_systemdependent.c.
Change-Id: I7b9738a7113c0c4687e5d320581ff69d98a8b271
Dmitry Kovalev [Thu, 27 Feb 2014 01:07:06 +0000 (17:07 -0800)]
Replacing int_mv with MV in single_motion_search().
Change-Id: I527c42fa776c772c26c027a47f3dada129f5c33e
Alex Converse [Thu, 27 Feb 2014 00:46:21 +0000 (16:46 -0800)]
Don't let GOLDEN_FRAME turn off LAST_FRAME as a reference when GOLDEN_FRAME is disabled.
There was a bug in the previous code that if GOLDEN was better than
LAST neither would be used. LAST would get turned off due to superior
GOLDEN quality then all GOLDEN modes would get skipped.
Change-Id: I173f3720451707dab7b2cbbe8b8e6a047089bde7
Ivan Maltz [Thu, 27 Feb 2014 00:36:23 +0000 (16:36 -0800)]
Merge "log overall psnr, average psnr, and overall mse"
Yaowu Xu [Thu, 27 Feb 2014 00:23:49 +0000 (16:23 -0800)]
Merge "Reduce the frequency of partition size adjsutment"
Dmitry Kovalev [Wed, 26 Feb 2014 23:01:30 +0000 (15:01 -0800)]
Adding vpx_sse_to_psnr() function.
Removing all copies of identical vp8_mse2psnr/vp9_mse2psnr functions.
Using vpx_sse_to_psnr() instead in all places.
Change-Id: I15beef9834d43d8fc8a8a7a2d1fc5de3d658fed8
Dmitry Kovalev [Thu, 27 Feb 2014 00:20:17 +0000 (16:20 -0800)]
Moving common code into vp9_get_entropy_contexts().
Change-Id: I197670a405f7c2e7796424faa187a61081db8567
Dmitry Kovalev [Thu, 27 Feb 2014 00:03:19 +0000 (16:03 -0800)]
Merge "Cleaning up vp9_encodemb.c file."
Dmitry Kovalev [Wed, 26 Feb 2014 22:56:00 +0000 (14:56 -0800)]
Cleaning up cost_coeffs().
Change-Id: I02f097fd7c4c787035f03114ddde5b04b2640b39
Marco Paniconi [Wed, 26 Feb 2014 22:35:14 +0000 (14:35 -0800)]
Merge "Update some comments in vpx_temporal_scalable_patterns."
Marco Paniconi [Wed, 26 Feb 2014 22:00:43 +0000 (14:00 -0800)]
Update some comments in vpx_temporal_scalable_patterns.
Change-Id: I747d26c544cc56feaaf7c75403f2f4c16a496bb7
Ivan Maltz [Wed, 26 Feb 2014 21:27:01 +0000 (13:27 -0800)]
log overall psnr, average psnr, and overall mse
Change-Id: I0b4fa1f8104b7fa5780d4e99e5401c14271a992d
Dmitry Kovalev [Wed, 26 Feb 2014 21:51:12 +0000 (13:51 -0800)]
Cleaning up vp9_encodemb.c file.
Usage of encode_b_args is unnecessary because encode_block_pass1() doesn't
use them. That's why optimize_init_b() call is also not required.
Change-Id: Ib6cfe4916c2ca85749c90bb0adcba6fea592f9ac
Johann [Wed, 26 Feb 2014 21:44:28 +0000 (13:44 -0800)]
Merge "VP8 for ARMv8 by using NEON intrinsics 05"
Minghai Shang [Wed, 26 Feb 2014 21:30:50 +0000 (13:30 -0800)]
[svc] Add target bitrate settings for each layers.
Change-Id: Ia7677fb436667bc4f76db71f65e4784f433f7826
Dmitry Kovalev [Wed, 26 Feb 2014 21:25:10 +0000 (13:25 -0800)]
Merge "Cleaning up vp9_rdopt.c."
Yaowu Xu [Wed, 26 Feb 2014 21:24:51 +0000 (13:24 -0800)]
Merge "Separate super_block_yrd() into two functions"
Johann [Wed, 26 Feb 2014 20:28:19 +0000 (12:28 -0800)]
Merge "VP8 for ARMv8 by using NEON intrinsics 04"
Johann [Wed, 26 Feb 2014 20:27:36 +0000 (12:27 -0800)]
Merge "VP8 for ARMv8 by using NEON intrinsics 03"
hkuang [Wed, 26 Feb 2014 19:56:45 +0000 (11:56 -0800)]
Merge "Fix a bug in intra prediction due to change in
25e55526301eba7d6e5c68e25402e9b2102976d8 ."
Minghai Shang [Tue, 25 Feb 2014 23:32:20 +0000 (15:32 -0800)]
Change for adding QP settings for key frames
Change-Id: I4dcabb60cb1185eb9a2efa18b50f17af272d2cd6
Jingning Han [Wed, 26 Feb 2014 17:52:23 +0000 (09:52 -0800)]
Run sub-pixel motion search for NEWMV mode test
As Yunqing suggested, this commit makes non-RD mode decision always run
sub-pixel motion search in NEWMV mode. The compression performance
gains becomes fairly significant after we enabled sub-pixel accuracy
motion compensated prediction to calculate SAD cost.
For test sequences pedestrian_area at 1080p and vidyo1 at 720p, the
runtime goes slower by 5%. For rtc test set, the compression performance
is improved by 21.20%.
Change-Id: I38cbfdd5c53d79423e1fafb3154f8ddeed63bbf0
Tom Finegan [Wed, 26 Feb 2014 18:42:44 +0000 (10:42 -0800)]
Merge "cq_test.cc: Use size_t for file size and unsigned int for target bitrate."
Yaowu Xu [Tue, 25 Feb 2014 22:15:00 +0000 (14:15 -0800)]
Reduce the frequency of partition size adjsutment
The commit change to use partitions sizes directly from last frame
for frames directly where last frame selects partitions sizes based
on coding efficiency.
On --rt --cpu-used=-5, the change hurts compression by 4% but reduces
encoding time by ~20%
Change-Id: Ia68665e5c8489b7bfcf5fac7768332fba88928e6
Yaowu Xu [Tue, 25 Feb 2014 19:59:56 +0000 (11:59 -0800)]
Separate super_block_yrd() into two functions
for inter and intra respectively
Change-Id: I7764d3af780072f7bd7bde201f454dcdd2816e76
Marco Paniconi [Wed, 26 Feb 2014 17:27:17 +0000 (09:27 -0800)]
Merge "Add consective frame loss to error_resilience test."
Jingning Han [Wed, 26 Feb 2014 17:20:24 +0000 (09:20 -0800)]
Merge "Use sub-pixel accuracy prediction non-RD mode"
Dmitry Kovalev [Wed, 26 Feb 2014 17:20:14 +0000 (09:20 -0800)]
Merge "Using clamp() for active_{best, worst}_quality."
Jingning Han [Wed, 26 Feb 2014 17:20:13 +0000 (09:20 -0800)]
Merge "Use dedicated variables to store the best mode"
Dmitry Kovalev [Wed, 26 Feb 2014 17:19:54 +0000 (09:19 -0800)]
Merge "Removing redundant vp9_clear_system_state() call."
Dmitry Kovalev [Wed, 26 Feb 2014 17:19:28 +0000 (09:19 -0800)]
Merge "Removing vp9_optimize_b() function."
Jim Bankoski [Wed, 26 Feb 2014 15:46:56 +0000 (07:46 -0800)]
remove ioc problem with rdcost calculation
Change-Id: Idd800fae21508699e9d973134817493763a65042
James Zern [Wed, 26 Feb 2014 07:11:49 +0000 (23:11 -0800)]
dct16x16_test: add NEON functions
note not all functions have NEON implementations:
- fdct/fht/iht
Change-Id: I6bb7ffe89b8cc23b642e19caf22f04f2d5e39087
Marco Paniconi [Tue, 25 Feb 2014 02:14:50 +0000 (18:14 -0800)]
Add consective frame loss to error_resilience test.
Modify existing test to also check the case of dropping
(i.e., skip decoding) a consecutive list of frames.
Change-Id: Ia8c1195559f952e86e6697996931d3a920c05ae3
James Yu [Tue, 17 Dec 2013 13:18:57 +0000 (21:18 +0800)]
VP8 for ARMv8 by using NEON intrinsics 05
Add dequantizeb_neon.c
- vp8_dequantize_b_loop_neon
vpxdec --summary --noblit ../videos/tears_of_steel_1080p.webm
Before => After, 13.25 => 13.23 (fps)
Change-Id: Iebe3b0c6ed2359c778b0570763c5681ae25fef0c
Signed-off-by: James Yu <james.yu@linaro.org>
Dmitry Kovalev [Wed, 26 Feb 2014 02:11:42 +0000 (18:11 -0800)]
Removing redundant vp9_clear_system_state() call.
This function is already called in vp9_initialize_rd_consts().
Change-Id: I88c0484a7274660287fa25fdf5cdb721d3284e01
Dmitry Kovalev [Wed, 26 Feb 2014 02:01:49 +0000 (18:01 -0800)]
Using clamp() for active_{best, worst}_quality.
Change-Id: Ia79f975f54829b8ae6b568ca5c05f4f1f541706f
James Yu [Tue, 17 Dec 2013 11:06:31 +0000 (19:06 +0800)]
VP8 for ARMv8 by using NEON intrinsics 04
Add dequant_idct_neon.c
- vp8_dequant_idct_add_neon
vpxdec --summary --noblit ../videos/tears_of_steel_1080p.webm
Before => After, 13.25 => 13.22 (fps)
Change-Id: Id48f39e1da58dd3d8d37658e94989411997f4f7c
Signed-off-by: James Yu <james.yu@linaro.org>
hkuang [Wed, 26 Feb 2014 01:54:33 +0000 (17:54 -0800)]
Fix a bug in intra prediction due to change in
25e55526301eba7d6e5c68e25402e9b2102976d8 .
Change-Id: I17ac67c3ced91ad4f057b296f7e8dc86a3389f26
Dmitry Kovalev [Wed, 26 Feb 2014 01:46:27 +0000 (17:46 -0800)]
Cleaning up vp9_rdopt.c.
Change-Id: Ic7553a037493107c33cfb9815178a54095a289ec
James Yu [Tue, 17 Dec 2013 11:03:16 +0000 (19:03 +0800)]
VP8 for ARMv8 by using NEON intrinsics 03
Add dc_only_idct_add_neon.c
- vp8_dc_only_idct_add_neon
vpxdec --summary --noblit ../videos/tears_of_steel_1080p.webm
Before => After, 13.25 => 13.24 (fps)
Change-Id: I5e9e277ec3a3ca67e13c8cc4c324a6fbe8a897fc
Signed-off-by: James Yu <james.yu@linaro.org>
Dmitry Kovalev [Wed, 26 Feb 2014 01:23:01 +0000 (17:23 -0800)]
Removing vp9_optimize_b() function.
Change-Id: If859c9d58a8fe3f478d409b9e9ca59bcee8fd35d
Jingning Han [Wed, 26 Feb 2014 00:12:35 +0000 (16:12 -0800)]
Use sub-pixel accuracy prediction non-RD mode
This commit builds the actual prediction block in sub-pixel accuracy
and uses which to calculate SAD for non-RD mode decision. In the trail
run on pedestrian_area at 1080p, rtc speed -7 runtime goes from
23495 ms -> 25107 ms (7% slower). The compression performance is
improved by 20.57% for rtc test set.
Change-Id: I438589cd103fe99f1b50c2d1939ac6ca43fa0157
Johann [Tue, 25 Feb 2014 22:53:55 +0000 (14:53 -0800)]
Merge "VP8 for ARMv8 by using NEON intrinsics 02"
Alex Converse [Tue, 25 Feb 2014 22:30:33 +0000 (14:30 -0800)]
Merge "Shrink dct_value_cost[] to int16_t."
Jingning Han [Tue, 25 Feb 2014 21:50:32 +0000 (13:50 -0800)]
Use dedicated variables to store the best mode
Use a set of dedicated variables to buffer the current best mode
in non-RD mode decision. This allows to use mode_info for more
complicated test in the non-RD process.
Change-Id: I6024c9feb0662afd3eb29f7017f7b5a5446f303f
Deb Mukherjee [Tue, 25 Feb 2014 21:46:21 +0000 (13:46 -0800)]
Merge "Adds variance based fixed size partitioning"
Alex Converse [Tue, 25 Feb 2014 21:06:42 +0000 (13:06 -0800)]
Shrink dct_value_cost[] to int16_t.
The largest value is 13358.
Change-Id: I7a6b024a92b6250933d9ebc0cad066b966c96bd4
Deb Mukherjee [Mon, 24 Feb 2014 23:21:13 +0000 (15:21 -0800)]
Adds variance based fixed size partitioning
Adds a method for determining a fixed size partition based on
variance of a 64x64 SB. This method is added to rtc speed 6.
Also fixes a bug in rtc_use_partition() and includes some
refactoring related to partitioning search, and some cosmetics.
Currently compared to speed 5, the coding efficiency of speed 6
is -19% and that of speed 7 is -55%, in cbr mode.
Change-Id: I057e04125a8b765906bb7d4bf7a36d1e575de7c6
Dmitry Kovalev [Tue, 25 Feb 2014 19:06:05 +0000 (11:06 -0800)]
Merge "Using vp9_subtract_plane instead of vp9_subtract_{sb, sby, sbuv}."
Dmitry Kovalev [Tue, 25 Feb 2014 19:04:37 +0000 (11:04 -0800)]
Merge "Consistent names for reference_mode functions."
Dmitry Kovalev [Tue, 25 Feb 2014 18:51:17 +0000 (10:51 -0800)]
Merge "Changing vp9_full_search_sad{, x3, x8} signatures."
Dmitry Kovalev [Tue, 25 Feb 2014 18:44:19 +0000 (10:44 -0800)]
Merge "Cleaning up vp9_refining_search_sad() function."
Tom Finegan [Tue, 25 Feb 2014 17:44:31 +0000 (09:44 -0800)]
Merge "Change dx_time data type in vpxdec.c"
Adrian Grange [Tue, 25 Feb 2014 16:48:14 +0000 (08:48 -0800)]
Merge "Trap request to display non-existant frame"
Yunqing Wang [Tue, 25 Feb 2014 16:28:03 +0000 (08:28 -0800)]
Merge "Disable adaptive pred filter for non-split mode"
Yaowu Xu [Tue, 25 Feb 2014 16:22:49 +0000 (08:22 -0800)]
Merge changes I7e96d619,I9a7631d5
* changes:
normalize int64_t high value to INT64_MAX
resolve issue with arm code failing unit test
Jim Bankoski [Tue, 25 Feb 2014 15:29:48 +0000 (07:29 -0800)]
normalize int64_t high value to INT64_MAX
Change-Id: I7e96d6199f882d35357f6a9a08c04ad0af5dc26e
Jim Bankoski [Tue, 25 Feb 2014 01:34:27 +0000 (17:34 -0800)]
resolve issue with arm code failing unit test
The optimizer did something funny with the code around
line 1412. Before the call to encode_sb split_dist was
set properly but after it was adjusted and converted to
a negative.
https://code.google.com/p/webm/issues/detail?id=714
Change-Id: I9a7631d5325ade2dc28c1030653a23eecec8721b
James Yu [Sat, 22 Feb 2014 06:51:11 +0000 (14:51 +0800)]
Change dx_time data type in vpxdec.c
Change dx_time data type to int64_t to prevent
test time overflow when decoding long video.
Change-Id: I3dd5e324a246843e07e635fd25c50e71e385ed70
Signed-off-by: James Yu <james.yu@linaro.org>
Marco Paniconi [Mon, 24 Feb 2014 23:04:38 +0000 (15:04 -0800)]
vpx_temporal_scalable_patterns: fix to encoding_bitrate computation.
Change-Id: I56fcb82c4334c7591b765b55b3c2912778af6db5
Yunqing Wang [Mon, 24 Feb 2014 22:05:06 +0000 (14:05 -0800)]
Disable adaptive pred filter for non-split mode
If sf->disable_split_mask is DISABLE_ALL_SPLIT, disable
sf->adaptive_pred_interp_filter to avoid unnecessary operations.
Change-Id: Icb59174b2f4e9a3c3c16a696deb8018e5bd999eb
Deb Mukherjee [Mon, 24 Feb 2014 21:35:56 +0000 (13:35 -0800)]
Merge "Adds an intermediate speed level for rtc"
Deb Mukherjee [Mon, 24 Feb 2014 20:08:56 +0000 (12:08 -0800)]
Adds an intermediate speed level for rtc
Moves the existing speed 6 to speed 7 and adds an
intermediate level 6 which is roughly in between
speeds 6 and 7 in both speed and coding efficiency.
Also includes some minor fixes/adjustments.
Change-Id: I98befc4d82d750e79fe426c457c4a2571f6b6cc7
Adrian Grange [Mon, 24 Feb 2014 18:15:21 +0000 (10:15 -0800)]
Trap request to display non-existant frame
If show_existing_frame indicates that the decoder should
display an existing (previously decoded) frame, add a
check to make sure that the signaled buffer does contain
a valid decoded frame.
Change-Id: Iac8c686b321827414d69a3f2d0467566911bcba2
Yaowu Xu [Sat, 22 Feb 2014 20:41:06 +0000 (12:41 -0800)]
added clamp of segment loop filter level
for ABSDATA mode, so segment loop filter level always fall in valid
range for both Absolute and delta modes.
Change-Id: If90df3411479533dbdab63f8ae088d2f5dd174a9
Yaowu Xu [Mon, 24 Feb 2014 16:28:07 +0000 (08:28 -0800)]
Merge "Added clamp of qindex to valid range"
James Zern [Mon, 24 Feb 2014 00:08:00 +0000 (16:08 -0800)]
fix vs9 build after obj_int_extract changes
broken since:
1d3ca08 obj_int_extract.bat: Pass obj_int_extract.exe path to obj_int_extract.bat.
Change-Id: Ic3de9f8c2b604e3aa9b905e1bc0b578963e01a4a
Dmitry Kovalev [Sun, 23 Feb 2014 16:28:43 +0000 (08:28 -0800)]
Merge "Combining if-else branches from different conditions."
Dmitry Kovalev [Sun, 23 Feb 2014 16:25:40 +0000 (08:25 -0800)]
Merge "Cleaning up vp9_mvref_common.{h, c}."
James Yu [Tue, 17 Dec 2013 10:28:11 +0000 (18:28 +0800)]
VP8 for ARMv8 by using NEON intrinsics 02
Add copymem_neon.c
- vp8_copy_mem16x16_neon
- vp8_copy_mem8x8_neon
- vp8_copy_mem8x4_neon
vpxdec --summary --noblit ../videos/tears_of_steel_1080p.webm
Before => After, 13.25 => 13.25 (fps)
Change-Id: Ib956b5a20522ff57dc8a580bf0aef7b252bddba6
Signed-off-by: James Yu <james.yu@linaro.org>
Yaowu Xu [Sat, 22 Feb 2014 20:30:18 +0000 (12:30 -0800)]
Added clamp of qindex to valid range
The qindex for a segment was not clamped in ABSDATA mode, which may
cause invalid memory access if an ill-formed stream has a negative
value in ABSDATA mode. This commit added clamp to make sure qindex
for a segment always fall into valid range.
Change-Id: I0a74d00f4ef40aec7edaeca1d03c8645e23ab08c
Jingning Han [Sat, 22 Feb 2014 19:54:20 +0000 (11:54 -0800)]
Merge "Periodically update mode and mv costs"
James Zern [Sat, 22 Feb 2014 19:43:37 +0000 (11:43 -0800)]
Merge "move nestegg to third_party"